compiler_rt introduces a weak and hidden symbol compilerrt_abort_impl
into our pre-linked _all.o object, we have to handle it.
Fixes#4273.
R=iant, rsc, r
CC=golang-dev
https://golang.org/cl/6783050
Feature extraction was tested before, but not the final diffs.
This CL breaks function main into a smaller main + testable
compareAPI.
No functional changes.
R=golang-dev, adg
CC=golang-dev
https://golang.org/cl/6820057
On 6g/linux:
benchmark old ns/op new ns/op delta
BenchmarkFDCT 4606 4241 -7.92%
BenchmarkIDCT 4187 3923 -6.31%
BenchmarkDecodeBaseline 3154864 3170224 +0.49%
BenchmarkDecodeProgressive 4072812 4017132 -1.37%
BenchmarkEncode 39406920 34596760 -12.21%
Stack requirements before (from 'go tool 6g -S'):
(scan.go:37) TEXT (*decoder).processSOS+0(SB),$1352-32
(writer.go:448) TEXT (*encoder).writeSOS+0(SB),$5344-24
after:
(scan.go:37) TEXT (*decoder).processSOS+0(SB),$1064-32
(writer.go:448) TEXT (*encoder).writeSOS+0(SB),$2520-24
Also, in encoder.writeSOS, re-use the yBlock scratch buffer for Cb and
Cr. This reduces the stack requirement slightly, but also avoids an
unlucky coincidence where a BenchmarkEncode stack split lands between
encoder.writeByte and bufio.Writer.WriteByte, which occurs very often
during Huffman encoding and is otherwise disasterous for the final
benchmark number. FWIW, the yBlock re-use *without* the s/int/int32/
change does not have a noticable effect on the benchmarks.
R=r
CC=golang-dev, rsc
https://golang.org/cl/6823043
Fix the problem with no cookie handling when sending
other than GET or HEAD request through
(*Client) Do(*Request) (*Resposne, error).
https://code.google.com/p/go/issues/detail?id=3985
Adds a function (*Client) send(*Request) (*Reponse, error):
- sets cookies from CookieJar to request,
- sends request
- parses a reply cookies and updates CookieJar
Fixes#3985
R=bradfitz
CC=gobot, golang-dev
https://golang.org/cl/6653049
Previously we converted a time to UTC *and* serialized the timezone of
the original time. With this change, we serialize a UTCTime in the
original timezone.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6817048
- make sure dclcontext == PAUTO only in function bodies
- introduce PDISCARD to discard declarations in bodies of repeated imports
- skip printing initializing OAS'es in export mode, assuming they only occur after ODCL's
- remove ODCL and the initializing OAS from inl.c:ishairy
- fix confused use of ->typecheck in typecheckinl: it's about the ->inl, not about the fn.
- debuging aids: print ntype on ONAMEs too and -Emm instead of -Ell.
fixes#2812
R=rsc
CC=golang-dev
https://golang.org/cl/6800043
includes step 0: synthesize outparams, from 6600044
includes step 1,2: give outparams loopdepth 0 and verify unchanged results
generate esc:$mask tags, but still tie to sink if a param has mask != 0
from 6610054
adds final steps:
- have esccall generate n->escretval, a list of nodes the function results flow to
- use these in esccall and ORETURN/OAS2FUNC/and f(g())
- only tie parameters to sink if tag is absent, otherwise according to mask, tie them to escretval
R=rsc, bradfitz
CC=dave, gobot, golang-dev, iant, rsc
https://golang.org/cl/6741044
Before this patch the test would close the file descriptor but
not the os.File. When the os.File was GC'ed, the finalizer
would close the file descriptor again. That would cause
problems if the same file descriptor were returned by a later
call to open in another test.
On my system:
> GOGC=30 go test
--- FAIL: TestPassFD (0.04 seconds)
passfd_test.go:62: FileConn: dup: bad file descriptor
FAIL
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6776053
setAddr was showing up in profiles due to string concatenation construction the os.File name field. netFD.sysfile's Name() is never used, except in dup() so I believe it is safe to avoid this allocation.
R=mikioh.mikioh, rsc
CC=golang-dev
https://golang.org/cl/6742058
- Allow secondary values below the default value in second form. This is
to support before tags for secondary values, as used by Chinese.
- Eliminate collation elements that are guaranteed to be immaterial
after a weight increment.
R=r
CC=golang-dev
https://golang.org/cl/6739051
includes step 0: synthesize outparams, from 6600044
step 1: give outparams loopdepth 0 and verify unchanged results
step 2: generate esc:$mask tags, but still tie to sink if a param has mask != 0
next step: use in esccall (and ORETURN with implicit OAS2FUNC) to avoid tying to sink
R=rsc
CC=golang-dev
https://golang.org/cl/6610054
in typecheck and walk, conversion from OAS2RECV to OAS2
and to OSELRECV2 duplicated the ->rlist->n to ->right
thereby destroying the strict tree-ness of the AST (up to
ONAMES) of course. Several recursions in esc.c and inl.c
and probably elsewhere assume nodes of the tree aren't duplicated.
rather than defensively code around this, i'd rather assert
these cases away and fix their cause.
(this was tripped in 6741044)
R=rsc
CC=golang-dev
https://golang.org/cl/6750043
It's common to use the go list command in shell scripts, but
currently it's awkward to print a string slice from the Package
type in a way that's easily parseable by the shell. For example:
go list -f '{{range .Deps}}{{.}}
{{end}}'
(and even that prints an unwanted new line at the end|).
To make this easier, this CL adds a "join" function to the
format template.
go list -f '{{join .Deps "\n"}}'
R=rsc, dsymonds, minux.ma, remyoudompheng, r
CC=golang-dev
https://golang.org/cl/6680044
Previously, multi-byte characters were not allowed. Also certain single-byte
characters, such as '-', were disallowed.
Fixes#3813.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6641052
I just realized that there is no good place for adding
exposed function or method tests because server, unicast
and multicast_test.go do test complicated multiple test
objects, platform behaviros, protocol behaviors and API,
at the same time. Perhaps splitting them into per test
object might be better, so this CL provides tests focused
on API.
R=rsc
CC=gobot, golang-dev
https://golang.org/cl/6501057
Someone new to the language may not know the connection between ints and arrays, which was the only thing that the previous error told you anything about.
Fixes#4256.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6739048
PauseNs is a circular buffer of recent pause times, and the
most recent one is at [((NumGC-1)+256)%256].
Also fix comments cross-linking the Go and C definition of
various structs.
R=golang-dev, rsc, bradfitz
CC=golang-dev
https://golang.org/cl/6657047
many small writes to a network may be less efficient that a few
large writes.
This fixes net/http's TestClientWrites, broken by 6565056 that
introduced Writer.ReadFrom. That test implicitly assumed that
calling io.Copy on a *bufio.Writer wouldn't write to the
underlying network until the buffer was full.
R=dsymonds
CC=bradfitz, golang-dev, mchaten, mikioh.mikioh
https://golang.org/cl/6743044
This is part 2 of 2 for issue 4028.
benchmark old ns/op new ns/op delta
BenchmarkWriterCopyOptimal 53293 28326 -46.85%
BenchmarkWriterCopyUnoptimal 53757 30537 -43.19%
BenchmarkWriterCopyNoReadFrom 53192 36642 -31.11%
Fixes#4028.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6565056
The value of cosines are cached in a global array
instead of being recomputed each time.
The test was terribly slow on arm.
R=golang-dev, dave, nigeltao
CC=golang-dev
https://golang.org/cl/6733046
Since this patch changes the way complex literals are written
in export data, there are a few other glitches.
Fixes#4159.
R=golang-dev, rsc
CC=golang-dev, remy
https://golang.org/cl/6674047
A 4-bit window is convenient because 4 divides both 32 and 64,
therefore we never have a window spanning words of the exponent.
Additionaly, the benefit of a 5-bit window is only 2.6% at 1024-bits
and 3.3% at 2048-bits.
This code is still not constant time, however.
benchmark old ns/op new ns/op delta
BenchmarkRSA2048Decrypt 17108590 11180370 -34.65%
Benchmark3PrimeRSA2048Decrypt 13003720 7680390 -40.94%
R=gri
CC=golang-dev
https://golang.org/cl/6716048
Define the properties of the arguments better. In particular,
explain that the path is (sort of) relative to the argument to
Walk.
Fixes#4119.
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6721048
- always return 1 for y <= 0
- document that the sign of m is ignored
- protect against div-0 panics by treating
m == 0 the same way as m == nil
- added extra tests
Fixes#4239.
R=agl, remyoudompheng, agl
CC=golang-dev
https://golang.org/cl/6724046
The RFC doesn't actually have an opinion on whether this is a fatal or
warning level alert, but common practice suggests that it should be a
warning.
This involves rebasing most of the tests.
Fixes#3413.
R=golang-dev, shanemhansen, rsc
CC=golang-dev
https://golang.org/cl/6654050
This patch is enough to fix compilation of
exp/types tests but only passes a stripped down
version of the appripriate torture test.
Update #4207.
R=dave, nigeltao, rsc, golang-dev
CC=golang-dev
https://golang.org/cl/6621061
To be clear, this supports decoding the bytes on the wire into an
in-memory image. There is no API change: jpeg.Decode will still not
return until the entire image is decoded.
The code is obviously more complicated, and costs around 10% in
performance on baseline JPEGs. The processSOS code could be cleaned up a
bit, and maybe some of that loss can be reclaimed, but I'll leave that
for follow-up CLs, to keep the diff for this one as small as possible.
Before:
BenchmarkDecode 1000 2855637 ns/op 21.64 MB/s
After:
BenchmarkDecodeBaseline 500 3178960 ns/op 19.44 MB/s
BenchmarkDecodeProgressive 500 4082640 ns/op 15.14 MB/s
Fixes#3976.
The test data was generated by:
# Create intermediate files; cjpeg on Ubuntu 10.04 can't read PNG.
convert video-001.png video-001.bmp
convert video-005.gray.png video-005.gray.pgm
# Create new test files.
cjpeg -quality 100 -sample 1x1,1x1,1x1 -progressive video-001.bmp > video-001.progressive.jpeg
cjpeg -quality 50 -sample 2x2,1x1,1x1 video-001.bmp > video-001.q50.420.jpeg
cjpeg -quality 50 -sample 2x1,1x1,1x1 video-001.bmp > video-001.q50.422.jpeg
cjpeg -quality 50 -sample 1x1,1x1,1x1 video-001.bmp > video-001.q50.444.jpeg
cjpeg -quality 50 -sample 2x2,1x1,1x1 -progressive video-001.bmp > video-001.q50.420.progressive.jpeg
cjpeg -quality 50 -sample 2x1,1x1,1x1 -progressive video-001.bmp > video-001.q50.422.progressive.jpeg
cjpeg -quality 50 -sample 1x1,1x1,1x1 -progressive video-001.bmp > video-001.q50.444.progressive.jpeg
cjpeg -quality 50 video-005.gray.pgm > video-005.gray.q50.jpeg
cjpeg -quality 50 -progressive video-005.gray.pgm > video-005.gray.q50.progressive.jpeg
# Delete intermediate files.
rm video-001.bmp video-005.gray.pgm
R=r
CC=golang-dev
https://golang.org/cl/6684046
caller of ioutil.TempFile() can use f.Name() to get "pathname"
of the temporary file, instead of just the "name" of the file.
Also remove an out-of-date comment about random number state.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6649054
Valgrind Massif result when linking godoc:
On amd64:
old new -/+
mem_heap_B 185844612 175358047 -5.7%
mem_heap_extra_B 773404 773137 -0.0%
On 386/ARM:
old new -/+
mem_heap_B 141775701 131289941 -7.4%
mem_heap_extra_B 737011 736955 -0.0%
R=golang-dev, r, dave
CC=golang-dev
https://golang.org/cl/6655045
Better explanation of width for floating-point values.
Explain that scanning does not handle %#q etc.
Fixes#4202.
Fixes#4206.
R=golang-dev, adg, rsc, iant
CC=golang-dev
https://golang.org/cl/6620074
I was an idiot and was thinking that a small base didn't matter
because the exponentiation would quickly make the number the same size
as the modulus. But, of course, the small base continues to make
multiplications unrealistically cheap throughout the computation.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6649048
It was suggested that it's too easy to use crypto/hmac insecurely and
I think that has some merit. This change adds a Equal function to
make it obvious that MAC values should be compared in constant time.
R=rsc, max
CC=golang-dev
https://golang.org/cl/6632044
If source are not available, then the stack looks like:
stack_test.go:40: /tmp/gobuilder/linux-amd64-race-72b15c5d6f65/go/src/pkg/runtime/debug/bla-bla-bla/src/pkg/runtime/debug/stack_test.go:15 (0x43fb11)
stack_test.go:40: /tmp/gobuilder/linux-amd64-race-72b15c5d6f65/go/src/pkg/runtime/debug/bla-bla-bla/src/pkg/runtime/debug/stack_test.go:18 (0x43fb7a)
stack_test.go:40: /tmp/gobuilder/linux-amd64-race-72b15c5d6f65/go/src/pkg/runtime/debug/bla-bla-bla/src/pkg/runtime/debug/stack_test.go:37 (0x43fbf4)
stack_test.go:40: /tmp/gobuilder/linux-amd64-race-72b15c5d6f65/go/src/pkg/testing/bla-bla-bla/src/pkg/testing/testing.go:301 (0x43b5ba)
stack_test.go:40: /tmp/gobuilder/linux-amd64-race-72b15c5d6f65/go/src/pkg/runtime/bla-bla-bla/src/pkg/runtime/proc.c:276 (0x410670)
stack_test.go:40:
which is 6 lines.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6637060
The compiler is crashing on the following code:
type TypeID int
func (t *TypeID) encodeType(x int) (tt TypeID, err error) {
switch x {
case 0:
return t.encodeType(x * x)
}
return 0, nil
}
The pass marks "return struct" {tt TypeID, err error} as used,
and this causes internal check failure.
I've added the test to:
https://golang.org/cl/6525052/diff/7020/src/pkg/runtime/race/regression_test.go
R=golang-dev, minux.ma, rsc
CC=golang-dev
https://golang.org/cl/6611049
Also:
- Refactored testing framework to permit easier
idempotency testing.
- Applied gofmt -w src misc
This CL depends on CL 6639044 being applied first.
Formatting is not idempotent for all files: In those
files the comment position has changed (due to missing
precise location information) and/or the comment formatting
cannot/is not aware of independent code re-formatting.
In general it is very hard to make format idempotent when
running it in one pass only. Leaving that aside for now.
Fixes#1835.
R=r, rsc
CC=golang-dev
https://golang.org/cl/6624051
Preparation for forthcoming CL 6624051: Will make it
easier to see if/what changes are incurred by it.
The alignment changes in this CL are due to CL 6610051
(fix to alignment heuristic) where it appears that an
old version of gofmt was run (and thus the correct
alignment updates were not done).
R=r
CC=golang-dev
https://golang.org/cl/6639044
I typoed the code and tried to parse all the way to the end of the
message. Therefore it fails when NPN is not the last extension in the
ServerHello.
Fixes#4088.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6637052
1. correctly initialize .plt.got entries (point to the 1st entry)
2. add section .rel.plt (FreeBSD insists PLT relocs to be there)
3. put relocs of .got.plt into .rel.plt
4. set ELFOSABI_FREEBSD in ELF header
R=rsc
CC=golang-dev
https://golang.org/cl/6643050
This is a part of a bigger change that adds data race detection feature:
https://golang.org/cl/6456044
The purpose of this patch is to provide coarse-grained synchronization
between all Read() and Write() calls.
R=rsc, bradfitz, alex.brainman
CC=golang-dev
https://golang.org/cl/6610064
Check for specific, important misalignment in garbage collector.
Not a complete fix for issue 599 but an important workaround.
Update #599.
R=golang-dev, iant, dvyukov
CC=golang-dev
https://golang.org/cl/6641049
No longer assume that e_lfanew (in the IMAGE_DOS_HEADER strcuture) is always one byte. It is now regarded as a 4 byte uint32.
Fixes#4177.
R=golang-dev, alex.brainman, dave, minux.ma
CC=golang-dev
https://golang.org/cl/6587048
Updates setup-godoc-app.bash to produce a working godoc app
by substituting the go1.0.x go/... packages with those from tip.
R=gri
CC=golang-dev
https://golang.org/cl/6587080
- Changed Check signature to take function parameters for
more flexibility: Now a client can interrupt type checking
early (via panic in one the upcalls) once the desired
type information or number of errors is reached. Default
use is still simple.
- Cleaned up main typechecking loops. Now does not neglect
_ declarations anymore.
- Various other cleanups.
R=golang-dev, r, rsc
CC=golang-dev
https://golang.org/cl/6612049
This code has been reviewed before. The most significant
change is to check_test which now can handle more than
one error at the same error position (due to spurious
errors - should not happen in praxis once error handling
has been fine-tuned). This change makes check_test easier
to use during development.
R=rsc
CC=golang-dev
https://golang.org/cl/6584057
1. R_ARM_CALL can also be used to call a PLT entry
2. add support for R_ARM_PC24 and R_ARM_JUMP24
3. refactor, remove D_PLT32 in favor of D_CALL
Fixes#4006.
R=rsc, dave
CC=fullung, golang-dev
https://golang.org/cl/6622057
The protection against segfaults does not completely solve
crashes and breaks test/fixedbugs/bug365.go
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6615058
Also add call to GC() to make it easier to re-enable the test.
Update #4155.
When we have precise GC merged, re-enable this test.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6622058
Don't assume the test has a clean environment within /tmp.
Use an actual new tempdir for its tests.
Fixes FreeBSD build failure as seen at:
http://build.golang.org/log/396738676356d7fb6bab6eaf1b97cac820f8a90f
--- FAIL: TestMkdirAllWithSymlink (0.00 seconds)
path_test.go:178: Mkdir /tmp/dir: mkdir /tmp/dir: file exists
FAIL
FAIL os 1.091s
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6615057
t.Log("line 1\nline 2\nline 3")
Old output:
=== RUN TestLine3
--- PASS: TestLine3 (0.00 seconds)
testing_test.go:25: line 1
line 2
line 3
PASS
New output:
=== RUN TestLine3
--- PASS: TestLine3 (0.00 seconds)
testing_test.go:24: line 1
line 2
line 3
PASS
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6613069
decoder struct, inside the unmappedzero limit, to eliminate some
TESTB instructions in the inner decoding loop.
benchmark old ns/op new ns/op delta
BenchmarkDecode 2943204 2746360 -6.69%
R=r, dave
CC=golang-dev
https://golang.org/cl/6625058
The nil dereference in the next few lines doesn't seem
to cause a segmentation fault on Unix, but does seem
to halt the Go compiler.
The following is a test case:
>>>
package main
func mine(int b) int {
return b + 2
}
func main() {
mine()
c = mine()
}
<<<
Without this change only the following is caught:
typecheck.go:3: undefined: b
typecheck.go:4: undefined: b
with it, we catch all the errors:
typecheck.go:3: undefined: b
typecheck.go:4: undefined: b
typecheck.go:10: undefined: c
typecheck.go:10: cannot assign to c .
R=rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6542060
to be consistent with the fdct function, and to ease any future
idct rewrites in assembly.
The BenchmarkIDCT delta is obviously just an accounting change and not
a real saving, but it does give an indication of what proportion of
time was spent in the actual IDCT and what proportion was in shift and
clip. The idct time taken is now comparable to fdct.
The BenchmarkFDCT delta is an estimate of benchmark noise.
benchmark old ns/op new ns/op delta
BenchmarkFDCT 3842 3837 -0.13%
BenchmarkIDCT 5611 3478 -38.01%
BenchmarkDecodeRGBOpaque 2932785 2929751 -0.10%
R=r
CC=golang-dev
https://golang.org/cl/6625057
refer to opacity. Those references were copy/pasted from the
image/png encoding benchmarks, which cares whether or not the
source image is opaque, but the JPEG encoder does not care.
R=r
CC=golang-dev
https://golang.org/cl/6623052
elimination) in idct.go.
benchmark old ns/op new ns/op delta
BenchmarkIDCT 5649 5610 -0.69%
BenchmarkDecodeRGBOpaque 2948607 2941051 -0.26%
The "type block" declaration moved so that idct.go is compilable
as a stand-alone file: "go tool 6g -S idct.go" works.
R=r
CC=golang-dev
https://golang.org/cl/6619056
The profiler collects goroutine blocking information similar to Google Perf Tools.
You may see an example of the profile (converted to svg) attached to
http://code.google.com/p/go/issues/detail?id=3946
The public API changes are:
+pkg runtime, func BlockProfile([]BlockProfileRecord) (int, bool)
+pkg runtime, func SetBlockProfileRate(int)
+pkg runtime, method (*BlockProfileRecord) Stack() []uintptr
+pkg runtime, type BlockProfileRecord struct
+pkg runtime, type BlockProfileRecord struct, Count int64
+pkg runtime, type BlockProfileRecord struct, Cycles int64
+pkg runtime, type BlockProfileRecord struct, embedded StackRecord
R=rsc, dave, minux.ma, r
CC=gobot, golang-dev, r, remyoudompheng
https://golang.org/cl/6443115
The Go run-time assumes that all SSE floating-point exceptions
are masked so that Go programs are not broken by such invalid
operations. By default, the 64-bit version of the Plan 9 kernel
masks only some SSE floating-point exceptions. Here, we mask
them all on a per-thread basis.
R=rsc, rminnich, minux.ma
CC=golang-dev
https://golang.org/cl/6592056
The AST representation is already identical. Making the
code (nearly) identical in the parser reduces code size
and ensures that the ast.ValueSpec nodes have the same
values (specifically, iota). This in turn permits the
sharing of much of the respective code in the typechecker.
While at it: type functions work now, so use them.
R=r
CC=golang-dev
https://golang.org/cl/6624047
Code for parsing email addresses was already partially part of the public API with "func (Header) AddressList". This CL adds a trivial implementation for two public methods to parse address and lists from a string. With tests.
R=dsymonds
CC=golang-dev
https://golang.org/cl/5676067
Adds a DecryptBlock function which takes a password and a *pem.Block and
returns the decrypted DER bytes suitable for passing into other crypto/x509
functions.
R=golang-dev, agl, leterip
CC=golang-dev
https://golang.org/cl/6555052
They now show the correct name, the byte offset on the line, and context for the failed evaluation.
Before:
template: three:7: error calling index: index out of range: 5
After:
template: top:7:20: executing "three" at <index "hi" $>: error calling index: index out of range: 5
Here top is the template that was parsed to create the set, and the error appears with the action
starting at byte 20 of line 7 of "top", inside the template called "three", evaluating the expression
<index "hi" $>.
Also fix a bug in index: it didn't work on strings. Ouch.
Also fix bug in error for index: was showing type of index not slice.
The real previous error was:
template: three:7: error calling index: can't index item of type int
The html/template package's errors can be improved by building on this;
I'll do that in a separate pass.
Extends the API for text/template/parse but only by addition of a field and method. The
old API still works.
Fixes#3188.
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6576058
This is a backward-compatible API change.
Without the correct <- position information,
certain channel types have incorrect position
information.
R=iant, iant
CC=golang-dev
https://golang.org/cl/6585063
This is the first part of a bigger change that adds data race detection feature:
https://golang.org/cl/6456044
This change makes gc compiler instrument memory accesses when supplied with -b flag.
R=rsc, nigeltao, lvd
CC=golang-dev
https://golang.org/cl/6497074
This fixes some example code in the tar package documentation, which
first refers to tar.NewWriter and then to Header, which is inconsistent
because NewWriter and Header are both in the tar namespace.
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6595050
This change allows the Go build and install tools to
recognize Plan 9 i386 and amd64 binaries.
R=rsc, r, rminnich
CC=golang-dev
https://golang.org/cl/6575064
This change updates CL 6576057 for exceptional cases where
return values from Syscall/RawSyscall functions are used.
The system calls return 32-bit integers. With the recent change
in size of `int' in Go for amd64, the type conversion was not
catching `-1' return values. This change makes the conversion
explicitly `int32'.
R=rsc, r
CC=golang-dev
https://golang.org/cl/6590047
Also mention that ignoring second blank identifier of range is required by the spec in the code.
Fixes#4173.
R=daniel.morsing, remyoudompheng, r
CC=golang-dev
https://golang.org/cl/6594043
FormatSelections tries to call a nil function value if lw is nil
and the final entry in the selections array is non-nil. Luckily,
this doesn't actually happen in practice since godoc doesn't use
this combination (no line numbers, but with selections).
R=gri
CC=gobot, golang-dev
https://golang.org/cl/6488106
The previous implementation was a mess with invariants
maintained inconsistently. Essentially reimplemented
the package:
- used a circular list as internal representation for
significantly simpler implementation with fewer
special cases while maintaining the illusion of
a nil-terminated doubly linked list externally
- more precise documentation
- cleaned up and simplified tests, added test case
for issue 4103.
No changes to the API or documented semantics.
All this said, I would be in favor of removing
this package eventually. container/ring provides
a faster implementation and a simpler and more
powerful API.
Fixes#4103.
R=r
CC=golang-dev
https://golang.org/cl/6569072
The string searching is implemented separately so other functions
may make use of it in the future.
benchmark old ns/op new ns/op delta
BenchmarkSingleMaxSkipping 125889 2474 -98.03%
BenchmarkSingleLongSuffixFail 16252 1996 -87.72%
BenchmarkSingleMatch 260793 136266 -47.75%
benchmark old MB/s new MB/s speedup
BenchmarkSingleMaxSkipping 79.43 4041.57 50.88x
BenchmarkSingleLongSuffixFail 61.65 501.81 8.14x
BenchmarkSingleMatch 57.52 110.08 1.91x
R=nigeltao
CC=golang-dev
https://golang.org/cl/6545049
The system calls return 32-bit integers. With the recent change
in size of `int' in Go for amd64, the type conversion was not
catching `-1' return values. This change makes the conversion
explicitly `int32'.
R=rsc, rminnich, npe, r
CC=golang-dev
https://golang.org/cl/6576057
Silly and small but easy to be consistent.
To make it worthwhile, I eliminated an allocation when using
%x on a byte slice.
Fixes#4149.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6574046
In two cases, registers were allocated too early resulting
in exhausting of available registers when nesting these
operations.
The case of method calls was due to missing cases in igen,
which only makes calls but doesn't allocate a register for
the result.
The case of 8-bit multiplication was due to a wrong order
in register allocation when Ullman numbers were bigger on the
RHS.
Fixes#3907.
Fixes#4156.
R=rsc
CC=golang-dev, remy
https://golang.org/cl/6560054
Low hanging fruit optimization. Will remove an expensive copy if the range variable is an array.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6564052
This CL adds support for the these 7 new instructions to 6a/6l in
preparation of the upcoming CL for AES-NI accelerated crypto/aes:
AESENC, AESENCLAST, AESDEC, AESDECLAST, AESIMC, AESKEYGENASSIST,
and PSHUFD.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/5970055
Do not check compiler/linker timestamps for packages that are in the
$GOROOT. Avoids trying to rebuild non-writable standard packages when
timestamps have not been retained on the Go binaries.
Fixes#4106.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6533053
The previous version was created by an idiot. This time, Rog Peppe
wrote the text. Thanks, Rog.
(== doesn't work on slices in general, so it makes no sense to
talk about in the context of DeepEqual.)
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6566054
This code relies on some functions that are not yet in staging,
but it get's harder to keep all this in sync in a piece-meal
fashion.
R=rsc
CC=golang-dev
https://golang.org/cl/6492124
More pieces of the typechecker code:
- Operands are temporary objects representing an expressions's
type and value (for constants). An operand is the equivalent of
an "attribute" in attribute grammars except that it's not stored
but only passed around during type checking.
- Constant operations are implemented in const.go. Constants are
represented as bool (booleans), int64 and *big.Int (integers),
*big.Rat (floats), complex (complex numbers), and string (strings).
- Error reporting is consolidated in errors.go. Only the first
dozen of lines is new code, the rest of the file contains the
exprString and typeString functions formerly in two separate
files (which have been removed).
This is a replacement CL for 6492101 (which was created without
proper use of hg).
R=rsc, r
CC=golang-dev
https://golang.org/cl/6500114
file old_size new_size base@c1ce95068533
bin/go 14717392 6287824 5918236
this huge size difference is due to GC data for runtime.mheap
(NOPTR dataflag is not obeyed).
R=rsc, dave
CC=golang-dev
https://golang.org/cl/6547051
Remove VERSION, which was forcing version to 'devel'.
Old:
$ go version
go version devel
New:
$ go version
go version devel +0a3866d6cc6b Mon Sep 24 20:08:05 2012 -0400
The date and time (and time zone) is that of the most recent commit,
not the time of the build itself. With some effort we could normalize
the zone, but I don't think it's worth the effort (more C coding,
since Mercurial is unhelpful).
R=r, dsymonds
CC=golang-dev
https://golang.org/cl/6569049
The assembly offsets were converted mechanically using
code.google.com/p/rsc/cmd/asmlint. The instruction
changes were done by hand.
Fixes#2188.
R=iant, r, bradfitz, remyoudompheng
CC=golang-dev
https://golang.org/cl/6550058
Both methods allow to access the IP ancillary data through
socket control messages.
This CL is required for CL 6482044; go.net/ipv4: new package.
R=rsc, r, dave
CC=golang-dev
https://golang.org/cl/6426047
Session resumption saves a round trip and removes the need to perform
the public-key operations of a TLS handshake when both the client and
server support it (which is true of Firefox and Chrome, at least).
R=golang-dev, bradfitz, rsc
CC=golang-dev
https://golang.org/cl/6555051
Was not handling &x.y[0] and &x.y.z correctly where
y is an array or struct-valued field (not a pointer).
R=ken2
CC=golang-dev
https://golang.org/cl/6551059
This CL makes the compiler understand that the type of
the len or cap of a map, slice, or string is 'int', not 'int32'.
It does not change the meaning of int, but it should make
the eventual change of the meaning of int in 6g a bit smoother.
Update #2188.
R=ken, dave, remyoudompheng
CC=golang-dev
https://golang.org/cl/6542059
Use explicit IntSize constant instead of 4.
This CL does not change the meaning of int, but it should make
the eventual change of the meaning of int on amd64 a bit
smoother.
Update #2188.
R=ken, dave
CC=golang-dev
https://golang.org/cl/6554076
This CL makes the size of an int controlled by a variable
in cgo instead of hard-coding 4 (or 32 bits) in various places.
Update #2188.
R=iant, r, dave
CC=golang-dev
https://golang.org/cl/6548061
This CL makes the runtime understand that the type of
the len or cap of a map, slice, or string is 'int', not 'int32',
and it is also careful to distinguish between function arguments
and results of type 'int' vs type 'int32'.
In the runtime, the new typedefs 'intgo' and 'uintgo' refer
to Go int and uint. The C types int and uint continue to be
unavailable (cause intentional compile errors).
This CL does not change the meaning of int, but it should make
the eventual change of the meaning of int on amd64 a bit
smoother.
Update #2188.
R=iant, r, dave, remyoudompheng
CC=golang-dev
https://golang.org/cl/6551067
Using offsets from Tos is cumbersome and we've had problems
in the past. Since it's only being used to grab the PID, we'll just
get that from the default TLS instead.
R=rsc, rminnich, npe
CC=golang-dev
https://golang.org/cl/6543049
various implementation of collation. The tool provides commands for soring,
regressing one implementation against another, and benchmarking.
Currently it includes collation implementations for the Go collator, ICU,
and one using Darwin's CoreFoundation framework.
To avoid building this tool in the default build, the colcmp tag has been
added to all files. This allows other tools/colcmp in this directory (e.g. it may make
sense to move maketables here) to be put in this directory as well.
R=r, rsc, mpvl
CC=golang-dev
https://golang.org/cl/6496118
This CL fixes code that incorrectly assumes that int is 32 bits wide.
Specifically, the socketpair system call expects a pointer to a pair
of int32s, not a pair of ints. Fix this inside the wrappers without
changing the APIs.
Update #2188.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6552063
Change the grammar so that field access is a proper operator.
This introduces a new node, ChainNode, into the public (but
actually internal) API of text/template/parse. For
compatibility, we only use the new node type for the specific
construct, which was not parseable before. Therefore this
should be backward-compatible.
Before, .X.Y was a token in the lexer; this CL breaks it out
into .Y applied to .X. But for compatibility we mush them
back together before delivering. One day we might remove
that hack; it's the simple TODO in parse.go/operand.
This change also provides grammatical distinction between
f
and
(f)
which might permit function values later, but not now.
Fixes#3999.
R=golang-dev, dsymonds, gri, rsc, mikesamuel
CC=golang-dev
https://golang.org/cl/6494119
The code was trying to interpret a uint32 n as a signed quantity
and then store it in an int. For this, int(n) currently works, but when
int becomes 64 bits one must write int(int32(n)) instead, to get
the 32-bit sign extension.
Update #2188.
R=golang-dev, nigeltao
CC=golang-dev
https://golang.org/cl/6551068
After each line, ast.Print would do a zero-length write,
which would hit the boundary condition on Plan 9 when
reading over pipes (since message boundaries are
preserved). This change makes sure we only do positive-
length writes.
R=rsc, rminnich, dave, r
CC=golang-dev
https://golang.org/cl/6558046
Revision 63f7abcae015 introduced a bug caused by
code assuming registers started at X5, not X0.
Fixes#4138.
R=rsc
CC=golang-dev, remy
https://golang.org/cl/6558043
The width was not being set on the address, which meant
that the optimizer could not find variables that overlapped
with it and mark them as having had their address taken.
This let to the compiler believing variables had been set
but never used and then optimizing away the set.
Fixes#4129.
R=ken2
CC=golang-dev
https://golang.org/cl/6552059
*** There is an API change here: the introduction of the
LatinOffset int in the RangeTable struct. ***
* Avoid checking Latin range multiple times for non-Latin runes.
* Use linear search when it is faster than binary search.
go test -calibrate runs the calibration for where the linear/binary
crossover should be.
benchmark old MB/s new MB/s speedup
BenchmarkFields 36.27 41.43 1.14x
BenchmarkFieldsFunc 36.23 41.38 1.14x
The speedup here is evenly split between the linear scans
and the LatinOffset change. Both are about 1.07x.
R=r
CC=bradfitz, gobot, golang-dev
https://golang.org/cl/6526048
OpenBSD now requires ELF binaries to have a PT_NOTE that identifies
it as an OpenBSD binary. Refactor the existing NetBSD ELF signature
code and implement support for OpenBSD ELF signatures.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6489131
Previously we tried to parse email constraints as the maximum base
distance (which is unused and must be omitted according to RFC 5280)
because they share the same tag number. Rather than complicate the
code further, this CL just ignores the unused values.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6552044
Always process the DWARF info, even when the const value is determined
using the debug data block. This ensures that the injected enum is
removed and future loads of the same constant do not trigger
inconsistent definitions.
Add tests for issues 2470 and 4054.
Fixes#4054.
R=golang-dev, fullung, dave, rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6501101
In a range loop, the presence of a value implies the presence of a key.
However, the presence of a value as an *ast.Ident does not imply that
the key is also an *ast.Ident, thus leading to a panic any time the
two argument form is used where the key is not an identifier.
R=golang-dev, adg, r
CC=golang-dev
https://golang.org/cl/6540045
(see also issue 3961).
hexdump -c testdata/crlf.input
0000000 / * \r \n \t S o u r c e c o n t
0000010 a i n i n g C R / L F l i n
0000020 e e n d i n g s . \r \n \t T h e
0000030 g o f m t ' e d o u t p u t
0000040 m u s t o n l y h a v e
0000050 L F \r \n \t l i n e e n d i n g
0000060 s . \r \n * / \r \n p a c k a g e
0000070 m a i n \r \n \r \n f u n c m a i
0000080 n ( ) { \r \n \t / / l i n e
0000090 c o m m e n t \r \n \t p r i n t l
00000a0 n ( " h e l l o , w o r l d !
00000b0 " ) / / a n o t h e r l i
00000c0 n e c o m m e n t \r \n \t p r i
00000d0 n t l n ( ) \r \n } \r \n
00000db
hexdump -c testdata/crlf.golden
0000000 / * \n \t S o u r c e c o n t a
0000010 i n i n g C R / L F l i n e
0000020 e n d i n g s . \n \t T h e g
0000030 o f m t ' e d o u t p u t m
0000040 u s t o n l y h a v e L F
0000050 \n \t l i n e e n d i n g s . \n
0000060 * / \n p a c k a g e m a i n \n
0000070 \n f u n c m a i n ( ) { \n \t
0000080 / / l i n e c o m m e n t \n
0000090 \t p r i n t l n ( " h e l l o ,
00000a0 w o r l d ! " ) / / a n o
00000b0 t h e r l i n e c o m m e n
00000c0 t \n \t p r i n t l n ( ) \n } \n
00000cf
R=rsc
CC=golang-dev
https://golang.org/cl/6526052
we need to restore %PATH% before calling "dist banner", so that
it could suggest the user to add %GOROOT%/bin to %PATH% if necessary.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6525049
use a function to get stdout and stderr, instead of depending
on a specific libc implementation.
also make test/run.go replace \r\n by \n before comparing
output.
Fixes#2121.
Part of issue 1741.
R=alex.brainman, rsc, r, remyoudompheng
CC=golang-dev
https://golang.org/cl/5847068
We use pkg path instead of file name (which contains $WORK) in section symbols names.
R=golang-dev, fullung, rsc, iant
CC=golang-dev
https://golang.org/cl/6445085
Fixes writing of function parameter, result lists which
consist of multiple named or unnamed items with same type.
Fixes#4011.
R=golang-dev, bsiegert, bradfitz, rsc
CC=golang-dev
https://golang.org/cl/6475062
This is the second part of a bigger change that adds data race detection feature:
https://golang.org/cl/6456044
This change makes the linker emit dependency on runtime/race package when supplied with -b flag.
R=rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6488074
This is the first part of a bigger change that adds data race detection feature:
https://golang.org/cl/6456044
Adds -race flag to go command.
API change:
+pkg go/build, type Context struct, InstallTag string
R=rsc
CC=golang-dev
https://golang.org/cl/6488075
This is a part of a bigger change that adds data race detection feature:
https://golang.org/cl/6456044
This change breaks circular dependency between runtime/race and syscall packages.
R=rsc
CC=golang-dev
https://golang.org/cl/6498079
The change is a preparation for the new scheduler.
It introduces runtime.park() function,
that will atomically unlock the mutex and park the goroutine.
It will allow to remove the racy readyonstop flag
that is difficult to implement w/o the global scheduler mutex.
R=rsc, remyoudompheng, dave
CC=golang-dev
https://golang.org/cl/6501077
During interface compare, the operands will be evaluated twice. The operands might include function calls for conversion, so make them cheap before comparing them.
R=rsc
CC=golang-dev
https://golang.org/cl/6498133
Fixes#3456.
This proposal is a reformulation of CL 5987063. This CL resets
the default GOARM value to 6 and allows the use of the VFPv3
optimisation if GOARM=7. Binaries built with this CL in place
will abort if GOARM=7 was used and the target host does not
support VFPv3.
R=minux.ma, rsc, ajstarks
CC=golang-dev
https://golang.org/cl/6501099
In switches without an expression, the compiler would not convert the implicit true to an interface, causing codegen errors.
Fixes#3980.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6497147
This also fixes the semantics of some corner cases with the empty
match. TODOs for genericReplacer in the tests are fixed.
benchmark old ns/op new ns/op delta
BenchmarkGenericNoMatch 71395 3132 -95.61%
BenchmarkGenericMatch1 75610 20280 -73.18%
BenchmarkGenericMatch2 837995 86725 -89.65%
R=nigeltao, rsc
CC=golang-dev
https://golang.org/cl/6492076
Give the right name for errors, and add a test to check we're
getting the errors we expect.
Also fix an ordering bug (calling add after stopParse) that
caused a nil indirection rather than a helpful error.
Fixes#3280.
R=golang-dev, adg
CC=golang-dev
https://golang.org/cl/6520043
Command.Start could crash before if no fds were available
because a nil *os.File of /dev/null was added to the cleanup
list, which crashed before returning the proper error.
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6514043
When reading Go files, read through import block.
When reading non-Go files, read only leading comments.
R=nigeltao, adg, r
CC=golang-dev
https://golang.org/cl/6493068
instead of variables. Several reasons:
- Encourage users of the API to minimize the number of creations and reuse Collate objects.
- Don't rule out the possibility of using initialization code for collators. For some locales
it will be possible to have very compact representations that can be quickly expanded
into a proper table on demand.
Other changes:
- Change name of root* vars to main*, as the tables are shared between locales.
- Added Locales() method to get a list of supported locales.
R=r
CC=golang-dev
https://golang.org/cl/6498107
X509KeyPair wasn't really supposed to allow the certificate and
key to be in the same file, but it did work if you put the key
first. Since some HTTPS servers support loading keys and certs
like this, this change makes it work in either order.
Fixes#3986.
R=golang-dev, dave, rsc
CC=golang-dev
https://golang.org/cl/6499103
This is the easy part of issue 3528.
(What to do about "noescape" is the hard part, left open.)
Update #3528.
R=mikesamuel, r, dsymonds
CC=golang-dev
https://golang.org/cl/6493113
Right now we only have 32-bit ints so that's a no-op.
Took the opportunity to check for some other invalid values too.
Suggestions for additions or modifications welcome.
R=agl
CC=golang-dev
https://golang.org/cl/6493112
This fixes a problem with ELF tools thinking they know the
format of the symbol table, as we do not use any of the
standard formats for that table.
This change will probably annoy the Plan 9 users, but I
believe there are other incompatibilities already that mean
they have to use a Go-specific nm.
Fixes#3473.
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6500117
go tool dist env -w is supposed to print a Windows batch file.
Normally Windows will execute batch files without \r before \n,
but issue 3060 reports that if the file ends up containing paths
written in Chinese, Windows 7 cannot execute it without the \r.
So add the \r.
Fixes#3060.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6498120
SO_PEERCRED on unix domain socket will fill a Ucred struct,
thus linux needs a custom Getsockopt variant.
Fixes#3836.
R=golang-dev, rsc, iant
CC=golang-dev
https://golang.org/cl/6445104
Rename the first argument of CompareAndSwapT and AddT s/val/addr/
for consistency with LoadT and StoreT.
R=rsc, r, dvyukov
CC=golang-dev
https://golang.org/cl/6494112
It is enough to load directly the data word and the itab word
from memory, so we save a LEA instruction for each method call,
and allow elimination of some extra temporaries.
Update #1914.
R=daniel.morsing, rsc
CC=golang-dev, remy
https://golang.org/cl/6501110
Removes an extra LEAL/LEAQ instructions there and usually saves
a useless temporary in the idiom
if err := foo(); err != nil {...}
Generated code is also less involved:
MOVQ err+n(SP), AX
CMPQ AX, $0
(potentially CMPQ n(SP), $0) instead of
LEAQ err+n(SP), AX
CMPQ (AX), $0
Update #1914.
R=daniel.morsing, nigeltao, rsc
CC=golang-dev, remy
https://golang.org/cl/6493099
This verifies existing behavior. Some replacements are arguably wrong
(these are marked with TODO) but changing behavior is left for a
follow-up CL.
Also fix that BenchmarkGenericMatch wasn't actually matching anything.
R=rsc, eric.d.eisner
CC=bradfitz, golang-dev
https://golang.org/cl/6488110
First set of type checker files for review.
The primary concern here is the typechecker
API (types.go).
R=rsc, adonovan, r, rogpeppe
CC=golang-dev
https://golang.org/cl/6490089
Broke tests on 386.
««« original CL description
6l/8l: emit correct opcodes to F(SUB|DIV)R?D.
When the destination was not F0, 6l and 8l swapped FSUBD/FSUBRD and
FDIVD/FDIVRD.
R=golang-dev, dave, rsc
CC=golang-dev
https://golang.org/cl/6498092
»»»
R=golang-dev
CC=golang-dev
https://golang.org/cl/6492100
Also, clear Content-Type and Content-Length on Not Modified
responses before server.go strips them and spams the logs with
warnings.
R=rsc
CC=golang-dev
https://golang.org/cl/6503090
Refactored build + buildTrie into build + buildOrdering.
Note that since the tailoring code is not checked in yet, all tailorings are identical
to root. The table therefore should not and does not grow at this point.
R=r
CC=golang-dev
https://golang.org/cl/6500087
The main case where it happens is when evaluating &s[i] without
bounds checking, which usually happens during range loops (i=0).
This allows registerization of the corresponding variables,
saving 16 bytes of stack frame for each such range loop and a
LEAQ instruction.
R=golang-dev, rsc, dave
CC=golang-dev, remy
https://golang.org/cl/6497073
Fixes#3911.
Requires CL 6449127.
dfc@qnap:~$ ./runtime.test
runtime: this CPU has no floating point hardware, so it cannot run
this GOARM=7 binary. Recompile using GOARM=5.
R=rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6442109
When generating enums use the debug data section instead of the
DWARF debug info, if it is available in the ELF file. This allows
mkerrors.sh to work correctly on OpenBSD/386 and NetBSD/386.
Fixes#2470.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6495090
The parser depends on it but the client might not import it, so make sure it's there.
Fixes#4038.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6497094
Fixes#3862.
There were many areas where conn.err was being accessed
outside the mutex. This proposal moves the err value to
an embedded struct to make it more obvious when the error
value is being accessed.
As there are no Benchmark tests in this package I cannot
feel confident of the impact of this additional locking,
although most will be uncontended.
R=dvyukov, agl
CC=golang-dev
https://golang.org/cl/6497070
order.
JPEG images generated prior to this CL are still valid JPEGs, as the
quantization tables used are encoded in the wire format. Such JPEGs just
don't use the recommended quantization tables.
R=r, dsymonds, raph, adg
CC=golang-dev, tuom.larsen
https://golang.org/cl/6497083
The old code was a depth first graph traversal that could, under the
right conditions, end up re-exploring the same subgraphs multiple
times, once for each way to arrive at that subgraph at a given depth.
The new code uses a breadth first search to make sure that it only
visits each reachable embedded struct once.
Also add fast path for the trivial case.
benchmark old ns/op new ns/op delta
BenchmarkFieldByName1 1321 187 -85.84%
BenchmarkFieldByName2 6118 5186 -15.23%
BenchmarkFieldByName3 8218553 42112 -99.49%
R=gri, r
CC=golang-dev
https://golang.org/cl/6458090
All of them call `newFileFD' which must properly restore close-on-exec on
duplicated fds.
R=golang-dev, bradfitz, mikioh.mikioh
CC=golang-dev
https://golang.org/cl/6445081
Signal handlers are global resources but many language
environments (Go, C++ at Google, etc) assume they have sole
ownership of a particular handler. Signal handlers in
mixed-language applications must therefore be robust against
unexpected delivery of certain signals, such as SIGPROF.
The default Go signal handler runtime·sigtramp assumes that it
will never be called on a non-Go thread, but this assumption
is violated by when linking in C++ code that spawns threads.
Specifically, the handler asserts the thread has an associated
"m" (Go scheduler).
This CL is a very simple workaround: discard SIGPROF delivered to non-Go threads. runtime.badsignal(int32) now receives the signal number; if it returns without panicking (e.g. sig==SIGPROF) the signal is discarded.
I don't think there is any really satisfactory solution to the
problem of signal-based profiling in a mixed-language
application. It's not only the issue of handler clobbering,
but also that a C++ SIGPROF handler called in a Go thread
can't unwind the Go stack (and vice versa). The best we can
hope for is not crashing.
Note:
- I've ported this to all POSIX platforms, except ARM-linux which already ignores unexpected signals on m-less threads.
- I've avoided tail-calling runtime.badsignal because AFAICT the 6a/6l don't support it.
- I've avoided hoisting 'push sig' (common to both function calls) because it makes the code harder to read.
- Fixed an (apparently incorrect?) docstring.
R=iant, rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6498057
Added instructions for starting an http server
to the godoc header for this package. With the old
instructions, the example "go tool pprof..." commands
wouldn't work unless you happen to be running an http
server on port 6060 in your application.
R=golang-dev, minux.ma, adg, giacomo.tartari
CC=golang-dev
https://golang.org/cl/6483049
cgo[1-4].go, go1.go couldn't be tested now
(cgo[1-4].go can only be tested when cgo is enabled, go1.go
contain a list of filenames in the current directory)
R=golang-dev, alex.brainman, rsc
CC=golang-dev
https://golang.org/cl/6218048
Accomplished by synchronizing the formatting of conversion errors between typecheck.c and subr.c
Fixes#3984.
R=golang-dev, remyoudompheng, rsc
CC=golang-dev
https://golang.org/cl/6500064
There was mail on golang-nuts a few weeks ago
from someone who understood the message perfectly
and knew he had a cyclic dependency but assumed
that Go, like Python or Java, was supposed to handle it.
R=golang-dev, bradfitz, dave
CC=golang-dev
https://golang.org/cl/6488069
for both locale-specific exemplar characters and tailorings to
the collation table.
Some specifices:
- Moved stringSet to the beginning of the file and added some functionality
to parse command line files.
- openReader now modifies the input URL for localFiles to guarantee that
any http source listed in the generated file is indeed this source.
- Note that the implementation of the Tailoring API used by maketables.go
is not yet checked in. So for now adding tailorings are simply no-ops.
- The generated file of exemplar characters will be used somewhere else.
Here is a snippet of how the body of the generated file looks like:
type exemplarType int
const (
exCharacters exemplarType = iota
exContractions
exPunctuation
exAuxiliary
exCurrency
exIndex
exN
)
var exemplarCharacters = map[string][exN]string{
"af": [exN]string{
0: "a á â b c d e é è ê ë f g h i î ï j k l m n o ô ö p q r s t u û v w x y z",
3: "á à â ä ã æ ç é è ê ë í ì î ï ó ò ô ö ú ù û ü ý",
4: "a b c d e f g h i j k l m n o p q r s t u v w x y z",
},
...
}
R=r
CC=golang-dev
https://golang.org/cl/6501070
- Elements in the array are now sorted as a linked list. This makes it easier to
apply tailorings.
- Added code to sort entries by collation elements.
- Added logical reset points. This is used for tailoring relative to certain
properties, rather than characters.
NOTE: all code for type entry should now be in order.go. To keep the diffs for
this CL reasonable, though, the existing code is left in builder.go. I'll move
this in a separate CL.
R=r
CC=golang-dev
https://golang.org/cl/6493063
Fixes#3525.
PTRACE_SYSCALL behaves like PTRACE_CONT and can deliver
a signal to the process. Ideally PtraceSingleStep should
support the signal argument, but its interface is frozen
by Go1.
R=golang-dev, r, rsc
CC=golang-dev
https://golang.org/cl/6353051
Parses a time header value into a time.Time according to rfc2616 sec 3.3.
R=golang-dev, dave, rsc, r
CC=bradfitz, golang-dev
https://golang.org/cl/6344046
A crash happens in the first request in a connection
if "params" field is missing because c.req.Params is Nil.
Fixes#3848.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6446051
Reverts part of CL 6460082.
If a doc comment describes a type by explaining the
meaning of one instance of the type, a leading article
is fine and makes the text less awkward.
Compare:
// A dog is a kind of animal.
// Dog is a kind of animal.
R=golang-dev, dsymonds, dvyukov, r
CC=golang-dev
https://golang.org/cl/6494066
This set of changes extends the Plan 9 support
to include the AMD64 architecture and should
work on all versions of Plan 9.
R=golang-dev, rminnich, noah.evans, rsc, minux.ma, npe
CC=akskuma, golang-dev, jfflore, noah.evans
https://golang.org/cl/6479052
This fixes a spurious 'invalid recursive type' error, and stops the compiler from emitting errors on uses of the invalid type.
Fixes#3766.
R=golang-dev, dave, minux.ma, rsc
CC=golang-dev
https://golang.org/cl/6443100
The Handler method makes the ServeMux dispatch logic
available to wrappers that enforce additional constraints
on requests.
R=golang-dev, bradfitz, dsymonds
CC=golang-dev
https://golang.org/cl/6450165
In serve mux, if pattern contains a host name, pass only the path to
the redirect handler.
Add tests for serve mux redirections.
R=rsc
CC=bradfitz, gobot, golang-dev
https://golang.org/cl/6329045
Also rename Node.{Add,Remove} to Node.{AppendChild,RemoveChild} to
be consistent with the DOM.
benchmark old ns/op new ns/op delta
BenchmarkParser 4042040 3749618 -7.23%
benchmark old MB/s new MB/s speedup
BenchmarkParser 19.34 20.85 1.08x
BenchmarkParser mallocs per iteration is also:
10495 before / 7992 after
R=andybalholm, r, adg
CC=golang-dev
https://golang.org/cl/6495061
This is required by the spec to produce the replacement char.
The fix lies in lib9's rune code.
R=golang-dev, nigeltao, rsc
CC=golang-dev
https://golang.org/cl/6443109
Other than catching an error case that was missed before, this
CL introduces no changes to the template language or API.
For simplicity, templates use spaces as argument separators.
This means that spaces are significant: .x .y is not the same as .x.y.
In the existing code, these cases are discriminated by the lexer,
but that means for instance that (a b).x cannot be distinguished
from (a b) .x, which is lousy. Although that syntax is not
supported yet, we want to support it and this CL is a necessary
step.
This CL emits a "space" token (actually a run of spaces) from
the lexer so the parser can discriminate these cases. It therefore
fixes a couple of undisclosed bugs ("hi".x is now an error) but
doesn't otherwise change the language. Later CLs will amend
the grammar to make .X a proper operator.
There is one unpleasantness: With space a token, three-token
lookahead is now required when parsing variable declarations
to discriminate them from plain variable references. Otherwise
the change isn't bad.
The CL also moves the debugging print code out of the lexer
into the test, which is the only place it's needed or useful.
Step towards resolving issue 3999.
It still remains to move field chaining out of the lexer
and into the parser and make field access an operator.
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6492054
In the example "units" program for goyacc, the exchange rates were
reciprocals of the correct amounts. Turn them right-side-up
and update them to current figures.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6495053
Fixes#3892.
Swapping the order of the writers inside the MultiWriter ensures
the request will be written to buf before http.ReadRequest completes.
The fencedBuffer is not required to make the test pass on
any machine that I have access too, but as the buf is shared
across goroutines, I think it is necessary for correctness.
R=bradfitz, fullung, franciscossouza
CC=golang-dev
https://golang.org/cl/6483061
Use version 2 of the NetBSD signal ABI - both version 2 and version 3
are supported by the kernel, with near identical behaviour. However,
the netbsd32 compat code does not allow version 3 to be used, which
prevents Go netbsd/386 binaries from running in compat mode on a
NetBSD amd64 kernel. Switch to version 2 of the ABI, which is the
same version currently used by NetBSD's libc.
R=minux.ma
CC=golang-dev
https://golang.org/cl/6476068
This shouldn't be an error (see issue 3999), but until it's handled
correctly, treat it as one to avoid confusion. Without this CL,
(A).X parses as two arguments.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6473059
Based on work by Russ Cox. From his CL:
This is generally useful but especially helpful when trying
to use the built-in boolean operators. It lets you write:
{{if not (f 1)}} foo {{end}}
{{if and (f 1) (g 2)}} bar {{end}}
{{if or (f 1) (g 2)}} quux {{end}}
instead of
{{if f 1 | not}} foo {{end}}
{{if f 1}}{{if g 2}} bar {{end}}{{end}}
{{$do := 0}}{{if f 1}}{{$do := 1}}{{else if g 2}}{{$do := 1}}{{end}}{{if $do}} quux {{end}}
The result can be a bit LISPy but the benefit in expressiveness and readability
for such a small change justifies it.
I believe no changes are required to html/template.
Fixes#3276.
R=golang-dev, adg, rogpeppe, minux.ma
CC=golang-dev
https://golang.org/cl/6482056
In the regtest data, surrogates are assigned primary weights based on
the surrogate code point value. Go now converts surrogates to FFFD, however,
meaning that the primary weight is based on this code point instead.
This change drops tests with surrogates and lets the tests pass.
R=r
CC=golang-dev
https://golang.org/cl/6461100
The second parameter for sql.putConn() (err) is always nil. As a result bad
connections are reused, even if the driver returns an driver.ErrBadConn.
Unsing a pointer to err instead achievs the desired behavior.
See http://code.google.com/p/go/issues/detail?id=3777 for more details.
Fixes#3777.
R=golang-dev, dave, bradfitz, jameshuachow, BlakeSGentry
CC=golang-dev
https://golang.org/cl/6348069
CVTSS2SQ's rounding mode is controlled by the RC field of MXCSR;
as we specifically need truncate semantic, we should use CVTTSS2SQ.
Fixes#3804.
R=rsc, r
CC=golang-dev
https://golang.org/cl/6352079
To make it more compliant.
This won't affect the behavior of running on OABI-only kernels.
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6475044
YearDay provides the day in the year represented by a given time.Time
object. This value is normally computed as part of other date calculations,
but not exported.
Fixes#3932.
R=golang-dev, r, remyoudompheng
CC=golang-dev, rsc
https://golang.org/cl/6460069
This makes SNI "just work" for callers using the standard http.Client.
Since we now have a test that depends on the httptest.Server cert, change
the cert to be a CA (keeping all other fields the same).
R=bradfitz
CC=agl, dsymonds, gobot, golang-dev
https://golang.org/cl/6448154
NetBSD's built-in linker script for 'ld -r' does not provide a
SEARCH_DIR. As a result libgcc.a is not found when -lgcc is used.
Work around this by determining the path to libgcc (by invoking
gcc with the -print-libgcc-file-name option) and explicitly
referencing the resulting library.
R=golang-dev, iant, aram, lucio.dere, minux.ma
CC=golang-dev
https://golang.org/cl/6470044
Moves the defer (again).
Also, correct the example documentation to match.
R=r, robert.hencke, iant, dsymonds, bradfitz
CC=golang-dev
https://golang.org/cl/6458158
Fixes#3793.
Tested using GOMAXPROCS=81 which was able to trigger a panic
in TestStressSurpriseServerCloses continually on a Core i5.
R=fullung, bradfitz
CC=golang-dev
https://golang.org/cl/6445069
The last fix was wrong w.r.t C's operator precedence,
and it also failed to really skip the NONE relocation.
The offending R_386_NONE relocation is a absolute
relocation in section .eh_frame.
TBR=golang-dev
CC=golang-dev
https://golang.org/cl/6463058
If ParseUint is like ParseInt it should have the same parameter
names, thus rename b to to base as in ParseInt's documentation.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6453152
with table changes.
NOTE: there is no test for this, but 1) the code has now the same
control flow as scan in exp/locale/collate/contract.go, which is
tested and 2) Builder verifies the generated table so bugs in this
code are quickly and easily found (which is how this bug was discovered).
R=r
CC=golang-dev
https://golang.org/cl/6461082
The main table will need to get a slightly different collation table as the one
used by regtest, as the regtest is based on the standard UCA DUCET, while
the locale-specific tables are all based on a CLDR root table.
This change allows changing the table without affecting the regression test.
R=r
CC=golang-dev
https://golang.org/cl/6453089
If we aren't explicit about the KeyUsages, the verifier
will treat the certificate as a server certificate and require
it to have a ExtKeyUsageServerAuth key usage.
R=golang-dev
CC=golang-dev
https://golang.org/cl/6453148
- pprof is a perl script, so go command should invoke
perl instead of trying to run pprof directly;
- pprof should use "go tool nm" unconditionally on windows,
no one else can extract symbols from Go program;
- pprof should use "go tool nm" instead of "6nm".
Fixes#3879.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6445082
Before, pointers always appeared as 0x1234ABCD. This CL
keeps that as the default for %p and %v, but lets explicit
numeric verbs override the default.
Fixes#3936.
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6441152
Only affects some tests and none seem likely to be problematic, but let's fix them.
Fixes#3971.
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6463060
When manipulating the stack pointer use the UESP register instead
of the ESP register, since the UESP register is the one that gets
restored from the machine context. Fixes broken tests on netbsd/386.
R=golang-dev, minux.ma, r, bsiegert
CC=golang-dev
https://golang.org/cl/6465054
This CL adds a step to the build procedure for cgo programs. It uses 'ld -r'
to combine all gcc compiled object file and generate a relocatable object file
for our ld. Additionally, this linking step will combine some static linking
gcc library into the relocatable object file, so that we can use libgcc,
libmingwex and libmingw32 without problem.
Fixes#3261.
Fixes#1741.
Added a testcase for linking in libgcc.
TODO:
1. still need to fix the INDIRECT_SYMBOL_LOCAL problem on Darwin/386.
2. still need to enable the libgcc test on Linux/ARM, because 5l can't deal
with thumb libgcc.
Tested on Darwin/amd64, Darwin/386, FreeBSD/amd64, FreeBSD/386, Linux/amd64,
Linux/386, Linux/ARM, Windows/amd64, Windows/386
R=iant, rsc, bradfitz, coldredlemur
CC=golang-dev
https://golang.org/cl/5822049
Now that the parser passes all tests in the test suite,
it is no longer necessary to keep track of which tests
pass and which don't. So remove the testlogs directory
and the code that uses it.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6453124
This currently fails on NetBSD due to the cloned file descriptors
that result from opening /dev/urandom. Disable the additional checking
until this is investigated and properly fixed.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6443129
Disable the crash handler test on NetBSD until I can figure out why
it triggers failures in later tests.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6460090
All of the remaining tests that had as status of PARSE rather than PASS had
good reasons for not passing the render-and-reparse step: the correct parse tree is
badly formed, so when it is rendered out as HTML, the result doesn't parse into the
same tree. So add them to the list of tests where that step is skipped.
Also, I discovered that it is possible to end up with HTML elements (not just text)
inside a raw text element through reparenting. So change the rendering routines to
handle that situation as sensibly as possible (which still isn't very sensible, but
this is HTML5).
R=nigeltao
CC=golang-dev
https://golang.org/cl/6446137
Optimize some common cases.
benchmark old ns/op new ns/op delta
BenchmarkScanFile 718907 667960 -7.09%
benchmark old MB/s new MB/s speedup
BenchmarkScanFile 23.03 25.51 1.11x
R=r
CC=golang-dev
https://golang.org/cl/6454150
When generating replacement elements for an <isindex> tag, the old
addSyntheticElement method was producing the wrong nesting. Replace
it with parseImpliedToken.
Pass the one remaining test in the test suite.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6453114
If a tag doesn't have a closing '>', it isn't considered a tag;
it is just ignored and EOF is returned instead.
Pass one additional test in the test suite.
Change tokenizer tests to match correct behavior.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6454131
In HTML content, having a self-closing tag is a parse error unless
the tag would be self-closing anyway (like <img>). The only place a
self-closing tag actually makes a difference is in XML-based foreign
content.
Pass 1 additional test.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6450109
Also simplified parsing of interface
types since they can only contain
methods (and no embedded interfaces)
in the export data.
R=rsc
CC=golang-dev
https://golang.org/cl/6446084
The keyword reprents an untyped nil and is useful for
passing nil values to methods and functions. The
nil will be promoted to the appropriate type when
used; if a type cannot be assigned, an error results.
R=rsc, dsymonds
CC=golang-dev
https://golang.org/cl/6459056
Although I don't use PAX enabled ARM kernels, PAX
does have support for ARM, so we're better off add
PT_PAX_FLAGS now in case people use PAX kernels.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6453092
Surrogate halves are part of UTF-16 and should never appear in UTF-8.
(The rune that two combined halves represent in UTF-16 should
be encoded directly.)
Encoding: encode as RuneError.
Decoding: convert to RuneError, consume one byte.
This requires changing:
package unicode/utf8
runtime for range over string
Also added utf8.ValidRune and fixed bug in utf.RuneLen.
Fixes#3927.
R=golang-dev, rsc, bsiegert
CC=golang-dev
https://golang.org/cl/6458099
If a table contained whitespace, text nodes would not get foster parented
correctly.
Pass 1 additional test.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6459054
Depends on CL 6197045.
Result obtained on Core i7 620M, Darwin/amd64:
benchmark old ns/op new ns/op delta
BenchmarkComplex128DivNormal 57 28 -50.78%
BenchmarkComplex128DivNisNaN 49 15 -68.90%
BenchmarkComplex128DivDisNaN 49 15 -67.88%
BenchmarkComplex128DivNisInf 40 12 -68.50%
BenchmarkComplex128DivDisInf 33 13 -61.06%
Result obtained on Core i7 620M, Darwin/386:
benchmark old ns/op new ns/op delta
BenchmarkComplex128DivNormal 89 50 -44.05%
BenchmarkComplex128DivNisNaN 307 802 +161.24%
BenchmarkComplex128DivDisNaN 309 788 +155.02%
BenchmarkComplex128DivNisInf 278 237 -14.75%
BenchmarkComplex128DivDisInf 46 22 -52.46%
Result obtained on 700MHz OMAP4460, Linux/ARM:
benchmark old ns/op new ns/op delta
BenchmarkComplex128DivNormal 1557 465 -70.13%
BenchmarkComplex128DivNisNaN 1443 220 -84.75%
BenchmarkComplex128DivDisNaN 1481 218 -85.28%
BenchmarkComplex128DivNisInf 952 216 -77.31%
BenchmarkComplex128DivDisInf 861 231 -73.17%
The 386 version has a performance regression, but as we have
decided to use SSE2 instead of x87 FPU for 386 too (issue 3912),
I won't address this issue.
R=dsymonds, mchaten, iant, dave, mtj, rsc, r
CC=golang-dev
https://golang.org/cl/6024045
The <title> element was getting removed from the stack of open elements,
when its parent, the <head> element should have been removed instead.
Pass 2 additional tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6449101
This is an attempt to fix windows build. Will restore
once we know what the problem is.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6461046
our old choice is not working properly at least on VFPv2 in
ARM1136JF-S (it's not preserved across float64->float32 conversions).
Fixes#3745.
R=dave, rsc
CC=golang-dev
https://golang.org/cl/6344078
Enhances test/run.go to support testing other directories
Will enable stdio tests on Windows in a follow-up CL.
R=golang-dev, alex.brainman, rsc
CC=golang-dev
https://golang.org/cl/6220049
When an element (like <nobr> or <p>) was implicitly closed by another
start tag, it would keep foster parenting from working because
the check for what was on top of the stack of open elements was
in the wrong place.
Move the check to addChild.
Pass 2 additional tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6460045
Hold ForkLock during dup of fd + cloexec in the net pkg,
per the locking policy documented in syscall/exec_unix.go.
R=golang-dev, dsymonds, adg
CC=golang-dev
https://golang.org/cl/6457080
Since NUL usually terminates strings in underlying syscalls, allowing
it when converting string arguments is a security risk, especially
when dealing with filenames. For example, a program might reason that
filename like "/root/..\x00/" is a subdirectory or "/root/" and allow
access to it, while underlying syscall will treat "\x00" as an end of
that string and the actual filename will be "/root/..", which might
be unexpected. Returning EINVAL when string arguments have NUL in
them makes sure this attack vector is unusable.
R=golang-dev, r, bradfitz, fullung, rsc, minux.ma
CC=golang-dev
https://golang.org/cl/6458050
Too many people use it without reading what it does.
Those people want ReadBytes or ReadString.
Fixes#3906.
R=golang-dev, iant, r
CC=golang-dev
https://golang.org/cl/6442087
The improvement is obtained by eliminating the zero
initialization of a large structure that is only
needed when the fast path fails.
Also add a missing roundtrip test for float32s.
benchmark old ns/op new ns/op delta
BenchmarkAppendFloatDecimal 301 180 -40.20%
BenchmarkAppendFloat 486 388 -20.16%
BenchmarkAppendFloatExp 492 383 -22.15%
BenchmarkAppendFloatNegExp 478 370 -22.59%
BenchmarkAppendFloatBig 650 541 -16.77%
BenchmarkAppendFloat32Integer 308 180 -41.56%
BenchmarkAppendFloat32ExactFraction 449 333 -25.84%
BenchmarkAppendFloat32Point 494 390 -21.05%
BenchmarkAppendFloat32Exp 488 387 -20.70%
BenchmarkAppendFloat32NegExp 488 378 -22.54%
R=r, rsc
CC=golang-dev, remy
https://golang.org/cl/6346081
The content of an HTML <title> element is RCDATA, but the content of an SVG
<title> element is parsed as tags. Now the parser doesn't go into RCDATA
mode in foreign content.
Pass 4 additional tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6448111
I have C functions implemented in .syso file (rather than .so or inlined in .go file).
W/o this change the gcc invocation fails with undefined symbols.
R=minux.ma, rsc
CC=golang-dev
https://golang.org/cl/6352076
The compiler is incorrectly rejecting switches on arrays of
comparable types. It also doesn't catch incomparable structs
when typechecking the switch, leading to unreadable errors
during typechecking of the generated code.
Fixes#3894.
R=rsc
CC=gobot, golang-dev, r, remy
https://golang.org/cl/6442074
These aren't needed for scalar multiplication, but since we export a
generic Add function we should handle it.
This change also corrects two bugs in p224Contract that it turned up.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6458076
PAX systems are Linux systems that are more paranoid about memory permissions.
These flags tell them to relax when running Go binaries.
Fixes#47.
R=iant
CC=golang-dev
https://golang.org/cl/6326054
This removes some (non-short only) tests that involved connections to
the outside world and thus were flakey. This does remove some test
coverage of the root fetchers on non-Windows platforms, but the right
place for that is crypto/x509.
R=golang-dev, krautz, rsc
CC=golang-dev
https://golang.org/cl/6455085
We already support reading ECDSA certificates and this change adds
write support.
R=golang-dev, bradfitz, rsc
CC=golang-dev
https://golang.org/cl/6422046
for dealing with CLDR files:
- Add now taxes a list of indexes of colelems that are variables. Checking and
handling is now done by the Builder. VariableTop is now also properly generated
using the Build method.
- Introduced separate Builder, called Tailoring, for creating tailorings of root
table. This clearly separates the functionality for building a table based on
weights (the allkeys* files) versus tables based on LDML XML files.
- Tailorings are now added by two calls instead of one: SetAnchor and Insert.
This more closely reflects the structure of LDML side and simplifies the
implementation of both the client and library side. It also preserves
some information that is otherwise hard to recover for the Builder.
- Allow the LDML XML element extend to be passed to Insert. This simplifies
both client and library implementation.
R=r
CC=golang-dev
https://golang.org/cl/6454061
Don't print unexported struct fields; their values are
not accessible via reflection.
Fixes#3898.
Also:
- added support for arrays
- print empty maps, arrays, slices, structs on one line
for a denser output
- added respective test cases
R=r
CC=golang-dev
https://golang.org/cl/6454089
Process a package's object in a reproducible
order (rather then in map order) so that we
get error messages in reproducible order.
R=r
CC=golang-dev
https://golang.org/cl/6449076
The text inside <script> tags is not ordinary raw text; there are all sorts
of other complications. This CL implements those complications.
Pass 76 additional tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6443070
This is more in sync with the rest of the package;
for instance, we have functions (not methods) to
deref or find the underlying type of a Type.
In the process use a single bytes.Buffer to create
the string representation for a type rather than
the (occasional) string concatenation.
R=r
CC=golang-dev
https://golang.org/cl/6458057
If an end tag has an attribute that is a quoted string containing '>',
the tokenizer would end the tag prematurely. Now it reads the attributes
on end tags just as it does on start tags, but the high-level interface
still doesn't return them, because their presence is a parse error.
Pass 1 additional test.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6457060
The receive operator was given incorrect precedence
resulting in incorrect deletion of parentheses.
Fixes#3843.
R=rsc
CC=golang-dev, remy
https://golang.org/cl/6442049
LLVM-based gcc will place all-zero data in a zero-filled
section, but our debug/macho can't handle that.
Fixes#3821.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6444049
If the X-Forwarded-For header already exists on a request, we
should append our client's IP to it after a comma+space instead
of overwriting it.
Fixes#3846.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6448053
This syscall was inadvertently exported when fixing
Getrlimit/Setrlimit on 32-bit platforms.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6458051
The situation only affects diagnostics but is easy to fix.
When computing lineNumber, use the position of the last item
returned by nextItem rather than the current state of the lexer.
This is internal only and does not affect the API.
Fixes#3886.
R=golang-dev, iant
CC=golang-dev
https://golang.org/cl/6445061
They were previously ignored when deciding order and
detecting dependency loops.
Fixes#3824.
R=rsc, golang-dev
CC=golang-dev, remy
https://golang.org/cl/6455055
Should probably be fixed in the syscall package, either
additional or instead of this CL.
Fixes#3842
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6442061
This fixes a data race (usually just harmlessly updating
"GET" to "GET"), but also follows RFC 2616 Sec 5.1.1 which
says that the request method is case-sensitive.
Fixes#3881
R=golang-dev, rsc, dsymonds
CC=golang-dev
https://golang.org/cl/6446063
Darwin kernels have a bug in accept() where error result from
an internal call is not checked and socket is accepted instead
of ECONNABORTED error. However, such sockets have no sockaddr,
which results in EAFNOSUPPORT error from anyToSockaddr, making
Go http servers running on Mac OS X easily susceptible to
denial of service from simple port scans with nmap.
Fixes#3849.
R=golang-dev, adg, mikioh.mikioh
CC=golang-dev
https://golang.org/cl/6456045
- Allow handles into the trie for different locales. Multiple tables share the same
try to allow for reuse of blocks.
- Significantly improved memory footprint and reduced allocations of trieNodes.
This speeds up generation by about 30% and allows keeping trieNodes around
for multiple locales during generation.
- Renamed print method to fprint.
R=r
CC=golang-dev
https://golang.org/cl/6408052
pkg log already appends a linefeed to the log message,
so log/syslog doesn't need to append another.
R=golang-dev, bradfitz, r
CC=golang-dev
https://golang.org/cl/6441048
When a cgo program calls setuid, setgid, etc., the GNU/Linux
pthread library sends signal SIGSETXID to each thread to tell
it to update its UID info. If Go is permitted to intercept
the default SIGSETXID signal handler, the program will hang.
This patch tells the runtime package to not try to intercept
SIGSETXID on GNU/Linux. This will be odd if a Go program
wants to try to use that signal, but it means that cgo
programs that call setuid, etc., won't hang.
Fixes#3871.
R=rsc, r, minux.ma, bradfitz
CC=golang-dev
https://golang.org/cl/6455050
The go/ast comment for FuncType.Params says that the field may be nil.
Make sure the printer accepts such a value. The go/parser always sets
the field (to provide parenthesis position information), but a program
creating a Go AST from scatch may not.
Added corresponding test case.
Fixes#3870.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6448060
If NUL bytes occur inside certain elements, convert them to U+FFFD
replacement character.
Pass 1 additional test.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6452047
Fixed creation of Func's, taking IsVariadic from parameter list rather
than results.
Updated checkObj to process ast.Fun objects.
R=gri
CC=golang-dev
https://golang.org/cl/6402046
Declare crosscall2. Declare the functions passed to it as
returning void, rather than relying on implicit return type.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6432060
If the body of an HTML document contains text, the <frameset> tag is
ignored. But not if the text is only whitespace.
Pass 4 additional tests.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6442043
If the key is not present, return value of the type of the element
not the type of the key. Also fix a test that should have caught this case.
Fixes#3850.
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6405078
Add commentary to explain better what's going on, but the
code change is a simple one-line reversal to the previous
form.
R=rsc
CC=golang-dev
https://golang.org/cl/6428072
Don't unescape entities in attributes when they don't end with
a semicolon and they are followed by '=', a letter, or a digit.
Pass 6 more tests from the WebKit test suite, plus one that was
commented out in token_test.go.
R=nigeltao
CC=golang-dev
https://golang.org/cl/6405073
Surrogates are still admitted, but I have sent mail to golang-dev on that topic.
Fixes#3785.
R=golang-dev, rogpeppe, iant
CC=golang-dev
https://golang.org/cl/6398049
If a Client's CheckRedirect function returns an error, we
again return both a non-nil *Response and a non-nil error.
Fixes#3795
R=golang-dev, n13m3y3r
CC=golang-dev
https://golang.org/cl/6429044
By default, all are still run, but a particular test can be
selected with the new flags.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6395053
Some class of bugs (data races, runtime bugs) can be found
only with real parallelism.
Note that GOMAXPROCS=32 is somewhat different from go test -cpu=32,
this intentionally uses GOMAXPROCS to stress program bootstrap,
testing code, garbage collections, etc.
Package selection is mostly random.
R=golang-dev, dave, r
CC=golang-dev
https://golang.org/cl/6346070
They can generate huge amounts of memory, causing failure on
small machines. Also they can be very slow. So slow that one test
was commented out! We uncomment it and use a flag.
Fixes#3742.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6373044
- Changed the representation of colElem to support a few cases
for some languages not supported by the current format.
- Changed offsets for implicit primary values. This makes the
values both easier to read and debug (last 4 nibbles are identical to
implicit primary value) and also results in better packing.
- Fixed bug in weight conversion code that did not pop up yet by
sheer luck.
Note that tables.go also includes changes to the contraction trie
from CL 6346092.
R=r, mpvl
CC=golang-dev
https://golang.org/cl/6392060
which has a rather large contraction table. The value of the next state
offset now starts after the current block, instead of before. This is
slightly less efficient (on extra addition per state change), but gives
some extra range for the offsets.
Also introduced constants for final (0) and noIndex (0xFF).
tables.go is updated in a separate CL.
R=r
CC=golang-dev
https://golang.org/cl/6346092
Gob decoding reads a whole message into memory and then
copies it into a bytes.Buffer. For large messages this wastes
an entire copy of the message. In this CL, we use a staging
buffer to avoid the large temporary.
Update #2539
RSS drops to 775MB from 1GB.
Active memory drops to 858317048 from 1027878136,
essentially the size of one copy of the input file.
R=dsymonds, nigeltao
CC=golang-dev
https://golang.org/cl/6392057
Allows a client to pre-allocate buffer space that is known to be necessary,
avoiding expensive reallocations.
R=gri, gri, adg
CC=golang-dev
https://golang.org/cl/6392061
The error was caused by a call to implements() even when
the type switch variable was not an interface.
Fixes#3786.
R=golang-dev, r
CC=golang-dev, remy
https://golang.org/cl/6354102
The previous implementation used to have a O(n) recursion
depth for unbalanced inputs. A test is added to check that a
reasonable amount of bytes is allocated in this case.
Fixes#3807.
R=golang-dev, dsymonds, gri
CC=golang-dev, remy
https://golang.org/cl/6345075
Also adds a new test for GOMAXPROCS=16 explicitly, which now passes
reliably in a stress loop like:
$ go test -c
$ (while ./http.test -test.v -test.run=Concurrency; do echo pass; done ) 2>&1 | tee foo; less foo
(It used to fail very quickly and reliably on at least Linux/amd64)
Fixes#3793
R=golang-dev, adg, r
CC=golang-dev
https://golang.org/cl/6347061
OpenSSL requires that RSA signatures be exactly the same byte-length
as the modulus. Currently it'll reject ~1/256 of our signatures: those
that end up a byte shorter.
Fixes#3796.
R=golang-dev, edsrzf, r
CC=golang-dev
https://golang.org/cl/6352093
Previously, strings that didn't have an explicit ASN.1 string type
were taken to be ASN.1 PrintableStrings. This resulted in an error if
a unrepresentable charactor was included.
For compatibility reasons, I'm too afraid to switch the default string
type to UTF8String, but this patch causes untyped strings to become
UTF8Strings if they contain a charactor that's not valid in a
PrintableString.
Fixes#3791.
R=golang-dev, bradfitz, r, r
CC=golang-dev
https://golang.org/cl/6348074
Entrust have issued a root certificate that's not marked as valid for
signing certificates.
This results in Go programs failing to validate certificates that
chain up to this root (i.e. gateway.push.apple.com:2195).
Although this is clearly a mistake on Entrust's part, it seems that we
will have to bodge around it.
R=golang-dev, dave, r
CC=golang-dev
https://golang.org/cl/6346064
Ran 'double.pl' on the pkg tree to identify doubled words.
One change to an error string return in x509; the rest are in comments.
Thanks to Matt Jibson for the idea.
R=golang-dev, bsiegert
CC=golang-dev
https://golang.org/cl/6344089
Made the following changes:
- Export errprintf() from all three OS-specific modules
- Added errprintf() to a.h
- Moved errprintf() in windows.c under xprintf(), since they are so similar
- Replaced all instances of xprintf() with errprintf() where a vflag check is done
Fixes#3788.
R=golang-dev, alex.brainman
CC=golang-dev
https://golang.org/cl/6346056
Move panic/defer/recover-related stuff from proc.c/runtime.c to a new file panic.c.
No semantic changes.
proc.c is 1800+ LOC and is a bit difficult to work with.
R=golang-dev, dave, r
CC=golang-dev
https://golang.org/cl/6343071
The use of math.Pow for mere squaring can be extremely
slow on soft-float ARM. Even on systems with hardware
floating-point, a speedup in test duration is observed.
On amd64
Before: ok math/rand 2.009s
After: ok math/rand 0.340s
Fixes#3740.
R=dave, golang-dev, r, r
CC=golang-dev
https://golang.org/cl/6348061
The previous version was not handling an off_t (64 bit)
parameter on 32 bit i386 systems. This patch splits sendfile
into two implementations in their respective arch-specific files.
Tested on FreeBSD amd64 and i386.
R=bradfitz
CC=golang-dev
https://golang.org/cl/6356048
1. Rename 'g' and 'm' local vars to 'gp' and 'mp' (convention already used in some functions)
'g' and 'm' are global vars that mean current goroutine and current machine,
when they are shadowed by local vars, it's confusing, no ability to debug log both, etc.
2. White-space shuffling.
No semantic changes.
In preparation to bigger changes.
R=golang-dev, dave
CC=golang-dev
https://golang.org/cl/6355061
There may be further savings if convT2I can avoid the function call
if the cache is good and T is uintptr-shaped, a la convT2E, but that
will be a follow-up CL.
src/pkg/runtime:
benchmark old ns/op new ns/op delta
BenchmarkConvT2ISmall 43 15 -64.01%
BenchmarkConvT2IUintptr 45 14 -67.48%
BenchmarkConvT2ILarge 130 101 -22.31%
test/bench/go1:
benchmark old ns/op new ns/op delta
BenchmarkBinaryTree17 8588997000 8499058000 -1.05%
BenchmarkFannkuch11 5300392000 5358093000 +1.09%
BenchmarkGobDecode 30295580 31040190 +2.46%
BenchmarkGobEncode 18102070 17675650 -2.36%
BenchmarkGzip 774191400 771591400 -0.34%
BenchmarkGunzip 245915100 247464100 +0.63%
BenchmarkJSONEncode 123577000 121423050 -1.74%
BenchmarkJSONDecode 451969800 596256200 +31.92%
BenchmarkMandelbrot200 10060050 10072880 +0.13%
BenchmarkParse 10989840 11037710 +0.44%
BenchmarkRevcomp 1782666000 1716864000 -3.69%
BenchmarkTemplate 798286600 723234400 -9.40%
R=rsc, bradfitz, go.peter.90, daniel.morsing, dave, uriel
CC=golang-dev
https://golang.org/cl/6337058
The switch at the beginning of the function already ensures n > 1,
so testing for n < 2 is redundant.
R=golang-dev, gri
CC=golang-dev
https://golang.org/cl/6350051
Occasionally I see:
--- FAIL: TestAllocations-15 (0.00 seconds)
all_test.go:1575: 6 mallocs after 100 iterations
Tested:
$ go test -cpu=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20 reflect
R=golang-dev, dsymonds, r
CC=golang-dev
https://golang.org/cl/6354063
Generating env.bat using dist env -wp > env.bat failed silently
if case of an error, because the message was redirected to env.bat.
Verbose messages still go to stdout, causing problems, but that's
a seperate change.
Made errprintf() identical to xprintf(), except for the output handle.
Yes, it's duplicate code, but most of the function is unpacking
the argument list and preparing it for WriteFile(), which has to be
done anyway.
R=golang-dev, alex.brainman
CC=golang-dev
https://golang.org/cl/6343047
Number represents the actual JSON text,
preserving the precision and
formatting of the original input.
R=rsc, iant
CC=golang-dev
https://golang.org/cl/6202068
go/build section "Build Constraints", first paragraph said:
"... they must be appear near the top of the file ..."
fixed to:
"... they must appear near the top of the file ..."
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6329060
Removes an incorrect code comment and some superfluous variables.
The comment I removed says that struct fields which implement
Unmarshaler must be pointers, even if they're in an addressable
struct. That's not the case, and there's already a test in decode_test.go
that demonstrates as much.
Encoding/json has quite a few assignments of reflect.Values to extra
variables – things like "iv := v" when there's no need to make a copy. I
think these are left over from a previous version of the reflect API. If they
aren't wanted, I wouldn't mind going through the package and getting
rid of the rest of them.
R=rsc
CC=golang-dev
https://golang.org/cl/6318047
This is a new, not yet committed API.
- Changed NewCommentMap to be independent of
*File nodes and more symmetric with the
Filter and Comments methods.
- Implemented Update method for use in
AST modifications.
- Implemented String method for debugging
R=rsc
CC=golang-dev
https://golang.org/cl/6303086
A few performance improvements, but without the stack sorting
change to avoid allocating, which is instead waiting on better
escape analysis.
R=rsc
CC=golang-dev
https://golang.org/cl/6265047
The issue seems to not be triggered right now,
but I've seen the deadlock after some other legal
modifications to runtime.
So I think we are safer this way.
R=rsc, r
CC=golang-dev
https://golang.org/cl/6339051
This can only happen if the hash function we're using is getting
far more than it's fair share of collisions, but that has happened
to us repeatedly as we've expanded the allowed use cases for
hash tables (issue 1544, issue 2609, issue 2630, issue 2883, issue 3695).
Maybe this will help the next time we try something new.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6306083
pkg/math/all_test.go tests Atan (and therefore Asin and Acos) to a
relative accuracy of 4e-16, but the test vector misses values where
the old algorithm was in error by more than that. For example:
x newError oldError
0.414215746 1.41e-16 -4.24e-16
0.414216076 1.41e-16 -4.24e-16
0.414217632 1.41e-16 -4.24e-16
0.414218770 1.41e-16 -4.24e-16
0.414225466 0 -5.65e-16
0.414226244 1.41e-16 -4.24e-16
0.414228756 0 -5.65e-16
0.414235089 0 -5.65e-16
0.414237070 0 -5.65e-16
R=rsc, golang-dev
CC=golang-dev
https://golang.org/cl/6302093
Several of my students were confused by trying to use both the error
return and a reply return, so I figured it was worth explicitly clarifying
that returning an error overrides the reply.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6327051
Flame motivated me to get around to adding extended key usage support
so that code signing certificates can't be used for TLS server
authentication and vice versa.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6304065
if we were to use sizeof(sa.sa_mask) instead of 8 as the last argument
to rt_sigaction, we would have already fixed this bug, so also updated
Linux/386 and Linux/amd64 files to use that; also test the return value
of rt_sigaction.
R=dave, rsc
CC=golang-dev
https://golang.org/cl/6297087
If the server replies with an HTTP response before we're done
writing our body (for instance "401 Unauthorized" response), we
were previously ignoring that, since we returned our write
error ("broken pipe", etc) before ever reading the response.
Now we read and write at the same time.
Fixes#3595
R=rsc, adg
CC=golang-dev
https://golang.org/cl/6238043
Also, fixes one violation found during testing where both
response and error could be non-nil when a CheckRedirect test
failed. This is arguably a minor API (behavior, not
signature) change, but it wasn't documented either way and was
inconsistent & non-Go like. Any code depending on the old
behavior was wrong anyway.
R=adg, rsc
CC=golang-dev
https://golang.org/cl/6307088
A comment to that effect was introduced
with rev d332f4b9cef5 but the respective
code wasn't deleted.
R=golang-dev, dsymonds
CC=golang-dev
https://golang.org/cl/6304086
Fixes a situation where a nested bad type would still
permit the outer type to install a working engine, leading
to inconsistent behavior.
Fixes#3273.
R=bsiegert, rsc
CC=golang-dev
https://golang.org/cl/6294067
A comment map associates comments with AST nodes
and permits correct updating of the AST's comment
list when the AST is manipulated.
R=rsc
CC=golang-dev
https://golang.org/cl/6281044
This slipped in with the implementation of Getpid in CL 5909043.
I'd exclude that CL entirely but it is tangled up in the Win32finddata changes.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/6297065
Preserve old API by using correct struct in system call
and then copying the results, as we did for SetsockoptLinger.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6307065
This new error is the only API change in the current draft of
Go 1.0.2 CLs. I'd like to include the CL that introduced it,
because it replaces a mysterious 'internal error' with a
useful error message, but I don't want any API changes,
so unexport the error constant for now. It can be
re-exported for Go 1.1.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6294055
Fixes#3708.
The fix to allow 5{c,g,l} to compile under clang 3.1 broke cross
compilation on darwin using the Apple default compiler on 10.7.3.
This failure was introduced in 9b455eb64690.
This has been tested by cross compiling on darwin/amd64 to linux/arm using
* gcc version 4.2.1 (Based on Apple Inc. build 5658) (LLVM build 2336.1.00)
* clang version 3.1 (branches/release_31)
As well as on linux/arm using
* gcc version 4.6.3 (Ubuntu/Linaro 4.6.3-1ubuntu5)
* Ubuntu clang version 3.0-6ubuntu3 (tags/RELEASE_30/final) (based on LLVM 3.0)
* Debian clang version 3.1-4 (branches/release_31) (based on LLVM 3.1)
R=consalus, rsc
CC=golang-dev
https://golang.org/cl/6307058
The type declarations were being generated using
a range over a map, which meant that successive
runs produced different orders. This will make sure
successive runs produce the same files.
Fixes#3707.
R=golang-dev, bradfitz
CC=golang-dev
https://golang.org/cl/6300062
If there are mutually recursive functions, there is a cycle in
the dependency graph, so the order is actually dependency order
among the strongly connected components: mutually recursive
functions get put into the same batch and analyzed together.
(Until now the entire package was put in one batch.)
The non-recursive case (single function, maybe with some
closures inside) will be able to be more precise about inputs
that escape only back to outputs, but that is not implemented yet.
R=ken2
CC=golang-dev, lvd
https://golang.org/cl/6304050
CL 4313064 fixed its test case but did not address a
general enough problem:
type T1 struct { F *T2 }
type T2 T1
type T3 T2
could still end up copying the definition of T1 for T2
before T1 was done being evaluated, or T3 before T2
was done.
In order to propagate the updates correctly,
record a copy of an incomplete type for re-execution
once the type is completed. Roll back CL 4313064.
Fixes#3709.
R=ken2
CC=golang-dev, lstoakes
https://golang.org/cl/6301059
This is part 1 of a 2 part changelist. Part 2 contains the mechanical
change to parse.go to compare atoms (ints) instead of strings.
The overall effect of the two changes are:
benchmark old ns/op new ns/op delta
BenchmarkParser 4462274 4058254 -9.05%
BenchmarkRawLevelTokenizer 913202 912917 -0.03%
BenchmarkLowLevelTokenizer 1268626 1267836 -0.06%
BenchmarkHighLevelTokenizer 1947305 1968944 +1.11%
R=rsc
CC=andybalholm, golang-dev, r
https://golang.org/cl/6305053
The reordering speedup in CL 6245068 changed the semantics
of %#v by delaying the clearing of some flags. Restore the old
semantics and add a test.
Fixes#3706.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6302048
CL 6250075 removed AI_MASK mask on all BSD variants,
however FreeBSD's AI_MASK does not include AI_V4MAPPED
and AI_ALL, and its libc is strict about the ai_flags.
This will fix the FreeBSD builder.
R=golang-dev, rsc
CC=golang-dev
https://golang.org/cl/6305054
On netbsd/386, tv_sec is a 64-bit integer for both timeval and timespec.
Fix the time handling code so that it works correctly.
R=golang-dev, rsc, m4dh4tt3r
CC=golang-dev
https://golang.org/cl/6256056
Thanks to Dave Cheney for the magic words "comm page".
benchmark old ns/op new ns/op delta
BenchmarkNow 197 33 -83.05%
This should make profiling a little better on OS X.
The raw time saved is unlikely to matter: what likely matters
more is that it seems like OS X sends profiling signals on the
way out of system calls more often than it should; avoiding
the system call should increase the accuracy of cpu profiles.
The 386 version would be similar but needs to do different
math for CPU speeds less than 1 GHz. (Apparently Apple has
never shipped a 64-bit CPU with such a slow clock.)
R=golang-dev, bradfitz, dave, minux.ma, r
CC=golang-dev
https://golang.org/cl/6275056
amd64 was done in CL 6275056.
We don't attempt to handle machines with clock speeds
less than 1 GHz. Those will fall back to the system call.
benchmark old ns/op new ns/op delta
BenchmarkNow 364 38 -89.53%
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6307045
Using an int64 for a block size doesn't make
sense on 32bit platforms but extracts a performance
penalty dealing with double word quantities on Arm.
linux/arm
benchmark old ns/op new ns/op delta
BenchmarkGobDecode 155401600 144589300 -6.96%
BenchmarkGobEncode 72772220 62460940 -14.17%
BenchmarkGzip 5822632 2604797 -55.26%
BenchmarkGunzip 326321 151721 -53.51%
benchmark old MB/s new MB/s speedup
BenchmarkGobDecode 4.94 5.31 1.07x
BenchmarkGobEncode 10.55 12.29 1.16x
R=golang-dev, rsc, bradfitz
CC=golang-dev
https://golang.org/cl/6272047